1. Field of the Invention
The present invention relates in general to magnetic flowmeters and in particular to an magnetic flowmeter having an improved arrangement of electrodes and inner wall insulation of the flow tube.
2. The Prior Art
There are several types of flowmeters used to measure the flow rate of a fluid in pipe or tube. Mechanical displacement flowmeters include an element, such as a screw or wheel, which is moved by the liquid and in turn moves dials or other indicators. This type of flowmeter is typically used in residential water meters.
Magnetic flowmeters, on the other hand, expose a flowing fluid to a magnetic field thereby inducing a voltage potential across the fluid. The induced potential is perpendicular to both the direction of the magnetic field and the direction of the fluid flow. Methods and apparatus have been well established in the prior art for measuring the flow of electrically conductive liquids in closed conduits by passing them through a magnetic field and detecting the voltage induced in a pair of electrodes at right angles to both the flow direction and the magnetic field in accordance with Faraday's Law of Induction. When the flowing conductive liquid in the conduit intersects the magnetic field, an induced signal is transferred to the electrodes.
Faraday's Law, upon which the designs of all magnetic flowmeters are based, shows that there is a linear relationship between the average velocity of a conductive liquid moving through a uniform magnetic field and the voltage induced in a pair of electrodes at right angles to both the liquid and the field. It is stated as E=K.times.B.times.D.times.V, where E is the generated voltage, K is a constant of proportionality, B is the magnetic field strength, D is the distance separating the electrodes, and V is the average velocity of the conductive liquid.
A magnetic flowmeter typically has a flow tube, through which a fluid flows, that terminates at each end with a flange. The flanges are used to integrate the flowmeter into a piping system. Between the flanges is a flow sensing unit, which subjects the fluid to a magnetic field. The sensing unit also measures the voltage potential difference, induced in the fluid by the fluid flowing through the magnetic field, between two sense electrodes. A frame of reference--usually a ground or earth potential--is required before the measurement of potential difference between the sense electrodes can be made meaningful. In fact, improper grounding considerations have long been a major cause of field installation problems with conventional magnetic flowmeters.
The inner surface of the flow tube must be insulated from the fluid to prevent the tube from grounding the induced potential difference. Because the inner surface of the flow tube is insulated from the fluid, ground connections generally are made to the conductive liquid by either grounding at least one of the entrance and exit flanges, or by installing special "grounding rings" at one or both of these locations. Common practice is to use two ground connections, either at the flanges or with grounding rings, at both ends of the flow tube, to provide electrostatic symmetry both upstream and downstream from the sense electrodes.
However, grounding at the remote entrance and exit flanges does not always provide adequate reduction in induced spurious electrical noise signals, which in some cases can destroy the reliable operation of the flowmeter. This problem is caused primarily because the physical placement of the actual ground reference points contributes substantially to the reliability of the measurement. The electrical resistance of the liquids generally encountered is not very low, and a high resistance ground return path can pick up a substantial amount of spurious electrical noise sufficient to swamp or substantially degrade the desired signal from the sense electrodes. Indeed, in practical installations, the spurious signals generated by electrolytic interaction between the conductive liquids and the electrodes, as well as signals originating from other sources, often totally mask the desired flow signal so as to render a permanent magnet system totally inoperative.
To avoid the spurious noise problem, it has been necessary to use electromagnetic coils excited by a pulsed DC system to periodically switch the magnetic field on and off, rather than to employ permanent magnets and a steady state magnetic field. The switching of the magnetic field of the coils permits sampling of spurious noise signals during both the on and off periods, thereby enabling the cancellation of spurious noise from the desired flow signal via differential subtraction circuitry.
A major disadvantage of the pulsed, electromagnetic field coil system is the requirement of a substantial power supply to furnish the field excitation. Present day electronic circuitry is capable of furnishing the required amplification and signal conditioning of the flow signal into meaningful measurements by employing only micropower. As a result, well up to 99 percent of the power supply requirements for an electromagnetic flowmeter can be consumed by the field coil excitation system alone. Furthermore, pulsed flowmeters must alternate between sampling induced voltage and background voltage. Therefore, they cannot monitor flow continuously.
U.S. Pat. No. 4,722,231, issued to Tanaka et al., shows a ground rod, instead of grounding rings, mounted no closer to the axis of the flow tube than one of the excitation coils, to provide direct ground contact with the flowing liquid. Although the ground rod is mounted in the same circumferential cross sectional plane as the pair of sensing electrodes it does not lead to a balanced distribution of the measured electrostatic field since it is used at only one side of the pipe.
U.S. Pat. No. 2,766,621, issued to Raynsford et al., shows an arrangement whereby two ground electrodes are located in the same plane as the sense electrodes. However, both the ground electrodes are permanently grounded to the shields of the connecting cables to the sense electrodes. In addition, they are also grounded to the flow tube itself, which eliminates the possibility of employing a Kelvin ground method to eliminate ground loop currents that can become non-common mode signals incapable of being eliminated from interfering with the desired flow signal. This also permanently affixes their function as being ground electrodes only, and does not permit their being switched to function as sense electrodes.
U.S. Pat. No. 3,965,738, issued to Watanabe, shows a pulsed D.C. method for exciting electromagnetic field coils, wherein the flow signal, in addition to spurious electrical noise, is sensed when the field is on, and spurious electrical noise only is sensed when the field is off, the latter to provide a noise adaptive zero, reference to be canceled by subtraction, to provide a noise insensitive flow reading. This method can only be used with electromagnetic excitation since the field generated by permanent magnets cannot effectively be switched off during a noise sensing mode.
U.S. Pat. No. 4,325,261, issued to Freund, Jr. et al., shows a method for ensuring constancy of electromagnetic field excitation by employing a fixed reference voltage in comparison to a voltage indicative of the current flow through the field coils, to compensate for variations within the field coil inductance. This method indirectly infers a relative measurement of the gauss field rather than directly quantifying the value of that field and cannot be used at all if permanent magnets are employed.
U.S. Pat. No. 4,459,858, issued to Marsh, shows a flowmeter having an electromagnetic sensor probe consisting of an inductance coil in combination with a plurality of electrodes serving as sense and reference ground. This method shows no means for noise cancellation unless a pulsed D.C. method as described by Watanabe is employed, in which case permanent magnets could not be utilized.
One disadvantage found in typical prior art magnetic flowmeters is that they lack any way of discharging residual charge that builds up on the sensing electrodes caused by friction of the fluids passing over the electrodes. This problem can lead to inaccurate velocity readings as the residual charge influences the potential sensed across the electrodes.